arXiv subjects

Song Wang

Publications and source records attributed to Song Wang.

At least 451 records · Page 25Linked to original sources

arXiv2016

Chandra ACIS Survey of X-ray Point Sources in Nearby Galaxies. II. X-ray Luminosity Functions and Ultraluminous X-ray Sources

Based on the recently completed {\it Chandra}/ACIS survey of X-ray point sources in nearby galaxies, we study the X-ray luminosity functions (XLFs) for X-ray point sources in different types of galaxies and the statistical properties of ultraluminous X-ray sources (ULXs). Uniform procedures are developed to compute the detection threshold, to estimate the foreground/background contamination, and to calculate the XLFs for individual galaxies and groups of galaxies, resulting in an XLF library for 343 galaxies of different types. With the large number of surveyed galaxies, we have studied the XLFs and ULX properties across different host galaxy types, and confirm with good statistics that the XLF slope flattens from lenticular ($α\sim1.50\pm0.07$) to elliptical ($\sim1.21\pm0.02$), to spirals ($\sim0.80\pm0.02$), to peculiars ($\sim0.55\pm0.30$), and to irregulars ($\sim0.26\pm0.10$). The XLF break dividing the neutron star and black hole binaries is also confirmed, albeit at quite different break luminosities for different types of galaxies. A radial dependency is found for ellipticals, with a flatter XLF slope for sources located between $D_{25}$ and 2$D_{25}$, suggesting the XLF slopes in the outer region of early-type galaxies are dominated by low-mass X-ray binaries in globular clusters. This study shows that the ULX rate in early-type galaxies is $0.24\pm0.05$ ULXs per surveyed galaxy, on a $5σ$ confidence level. The XLF for ULXs in late-type galaxies extends smoothly until it drops abruptly around $4\times10^{40}$ erg s$^{-1}$, and this break may suggest a mild boundary between the stellar black hole population possibly including 30 $M_\odot$ black holes with super-Eddington radiation and intermediate mass black holes.

astro-ph.GA↗

arXiv2016

Who Leads the Clothing Fashion: Style, Color, or Texture? A Computational Study

It is well known that clothing fashion is a distinctive and often habitual trend in the style in which a person dresses. Clothing fashions are usually expressed with visual stimuli such as style, color, and texture. However, it is not clear which visual stimulus places higher/lower influence on the updating of clothing fashion. In this study, computer vision and machine learning techniques are employed to analyze the influence of different visual stimuli on clothing-fashion updates. Specifically, a classification-based model is proposed to quantify the influence of different visual stimuli, in which each visual stimulus's influence is quantified by its corresponding accuracy in fashion classification. Experimental results demonstrate that, on clothing-fashion updates, the style holds a higher influence than the color, and the color holds a higher influence than the texture.

cs.CV↗

arXiv2016

Sharp Chandra View of ROSAT All-Sky Survey Bright Sources: I. Improvement of Positional Accuracy

The ROSAT All-Sky Survey (RASS) represents one of the most complete and sensitive soft X-ray all-sky surveys to date. However, the deficient positional accuracy of the RASS Bright Source Catalog (BSC) and subsequent lack of firm optical identifications affect the multi-wavelength studies of X-ray sources. The widely used positional errors $σ_{pos}$ based on the Tycho Stars Catalog (Tycho-1) have previously been applied for identifying objects in the optical band. The considerably sharper Chandra view covers a fraction of RASS sources, whose $σ_{pos}$ could be improved by utilizing the sub-arcsec positional accuracy of Chandra observations. We cross-match X-ray objects between the BSC and \emph{Chandra} sources extracted from the Advanced CCD Imaging Spectrometer (ACIS) archival observations. A combined counterparts list (BSCxACIS) with \emph{Chandra} spatial positions weighted by the X-ray flux of multi-counterparts is employed to evaluate and improve the former identifications of BSC with the other surveys. Based on these identification evaluations, we suggest that the point-likeness of BSC sources and INS (isolated neutron stars) candidates should be carefully reconsidered.

astro-ph.HE↗

arXiv2016

Chandra ACIS Survey of X-ray Point Sources: The Source Catalog

The $Chandra$ archival data is a valuable resource for various studies on different topics of X-ray astronomy. In this paper, we utilize this wealth and present a uniformly processed data set, which can be used to address a wide range of scientific questions. The data analysis procedures are applied to 10,029 ACIS observations, which produces 363,530 source detections, belonging to 217,828 distinct X-ray sources. This number is twice the size of the $Chandra$ Source Catalog (Version 1.1). The catalogs in this paper provide abundant estimates of the detected X-ray source properties, including source positions, counts, colors, fluxes, luminosities, variability statistics, etc. Cross-correlation of these objects with galaxies shows 17,828 sources are located within the $D_{25}$ isophotes of 1110 galaxies, and 7504 sources are located between the $D_{25}$ and 2$D_{25}$ isophotes of 910 galaxies. Contamination analysis with the log$N$--log$S$ relation indicates that 51.3\% of objects within 2$D_{25}$ isophotes are truly relevant to galaxies, and the "net" source fraction increases to 58.9\%, 67.3\%, and 69.1\% for sources with luminosities above $10^{37}$, $10^{38}$, and $10^{39}$ erg s$^{-1}$. Among the possible scientific uses of this catalog, we discuss the possibility to study intra-observation variability, inter-observation variability, and supersoft sources.

astro-ph.SR↗

arXiv2016

On Study of the Binarized Deep Neural Network for Image Classification

Recently, the deep neural network (derived from the artificial neural network) has attracted many researchers' attention by its outstanding performance. However, since this network requires high-performance GPUs and large storage, it is very hard to use it on individual devices. In order to improve the deep neural network, many trials have been made by refining the network structure or training strategy. Unlike those trials, in this paper, we focused on the basic propagation function of the artificial neural network and proposed the binarized deep neural network. This network is a pure binary system, in which all the values and calculations are binarized. As a result, our network can save a lot of computational resource and storage. Therefore, it is possible to use it on various devices. Moreover, the experimental results proved the feasibility of the proposed network.

cs.NE↗

arXiv2015

Relativistic baryonic jets from an ultraluminous supersoft X-ray source

The formation of relativistic jets by an accreting compact object is one of the fundamental mysteries of astrophysics. While the theory is poorly understood, observations of relativistic jets from systems known as microquasars have led to a well-established phenomenology. Relativistic jets are not expected from sources with soft or supersoft X-ray spectra, although two such systems are known to produce relatively low-velocity bipolar outflows. Here we report optical spectra of an ultraluminous supersoft X-ray source (ULS) in the nearby galaxy M81 (M81 ULS-1) showing blueshifted broad Hα emission lines, characteristic of baryonic jets with relativistic speeds. The time variable jets have projected velocities ~17 per cent of the speed of light, and seem similar to those in the prototype microquasar SS 433. Such relativistic jets are not expected to be launched from white dwarfs, but an origin from a black hole or neutron star in M81 ULS-1 is hard to reconcile with its constant soft X-rays. The completely unexpected presence of relativistic jets in a ULS challenges the canonical theories for jet formation, but may possibly be explained by a long speculated super-critically accreting black hole with optically thick outflows

astro-ph.HE↗

arXiv2015

LooseCut: Interactive Image Segmentation with Loosely Bounded Boxes

One popular approach to interactively segment the foreground object of interest from an image is to annotate a bounding box that covers the foreground object. Then, a binary labeling is performed to achieve a refined segmentation. One major issue of the existing algorithms for such interactive image segmentation is their preference of an input bounding box that tightly encloses the foreground object. This increases the annotation burden, and prevents these algorithms from utilizing automatically detected bounding boxes. In this paper, we develop a new LooseCut algorithm that can handle cases where the input bounding box only loosely covers the foreground object. We propose a new Markov Random Fields (MRF) model for segmentation with loosely bounded boxes, including a global similarity constraint to better distinguish the foreground and background, and an additional energy term to encourage consistent labeling of similar-appearance pixels. This MRF model is then solved by an iterated max-flow algorithm. In the experiments, we evaluate LooseCut in three publicly-available image datasets, and compare its performance against several state-of-the-art interactive image segmentation algorithms. We also show that LooseCut can be used for enhancing the performance of unsupervised video segmentation and image saliency detection.

cs.CV↗

arXiv2015

Two Optical Counterpart Candidates of M82-X1 from HST Observations

Optical counterparts can provide significant constraints on the physical nature of ultraluminous X-ray sources (ULXs). In this letter, we identify six point sources in the error circle of a ULX in M82, namely M82 X-1, by registering Chandra positions onto Hubble Space Telescope images. Two objects are considered as optical counterpart candidates of M82 X-1, which show F658N flux excess compared to the optical continuum that may suggest the existence of an accretion disk. The spectral energy distributions of the two candidates match well with the spectra for supergiants, with stellar types as F5-G0 and B5-G0, respectively. Deep spatially resolved spectroscopic follow-up and detailed studies are needed to identify the true companion and confirm the properties of this BH system.

astro-ph.SR↗

arXiv2015

Co-interest Person Detection from Multiple Wearable Camera Videos

Wearable cameras, such as Google Glass and Go Pro, enable video data collection over larger areas and from different views. In this paper, we tackle a new problem of locating the co-interest person (CIP), i.e., the one who draws attention from most camera wearers, from temporally synchronized videos taken by multiple wearable cameras. Our basic idea is to exploit the motion patterns of people and use them to correlate the persons across different videos, instead of performing appearance-based matching as in traditional video co-segmentation/localization. This way, we can identify CIP even if a group of people with similar appearance are present in the view. More specifically, we detect a set of persons on each frame as the candidates of the CIP and then build a Conditional Random Field (CRF) model to select the one with consistent motion patterns in different videos and high spacial-temporal consistency in each video. We collect three sets of wearable-camera videos for testing the proposed algorithm. All the involved people have similar appearances in the collected videos and the experiments demonstrate the effectiveness of the proposed algorithm.

cs.CV↗

arXiv2015

Unsupervised Cross-Domain Recognition by Identifying Compact Joint Subspaces

This paper introduces a new method to solve the cross-domain recognition problem. Different from the traditional domain adaption methods which rely on a global domain shift for all classes between source and target domain, the proposed method is more flexible to capture individual class variations across domains. By adopting a natural and widely used assumption -- "the data samples from the same class should lay on a low-dimensional subspace, even if they come from different domains", the proposed method circumvents the limitation of the global domain shift, and solves the cross-domain recognition by finding the compact joint subspaces of source and target domain. Specifically, given labeled samples in source domain, we construct subspaces for each of the classes. Then we construct subspaces in the target domain, called anchor subspaces, by collecting unlabeled samples that are close to each other and highly likely all fall into the same class. The corresponding class label is then assigned by minimizing a cost function which reflects the overlap and topological structure consistency between subspaces across source and target domains, and within anchor subspaces, respectively.We further combine the anchor subspaces to corresponding source subspaces to construct the compact joint subspaces. Subsequently, one-vs-rest SVM classifiers are trained in the compact joint subspaces and applied to unlabeled data in the target domain. We evaluate the proposed method on two widely used datasets: object recognition dataset for computer vision tasks, and sentiment classification dataset for natural language processing tasks. Comparison results demonstrate that the proposed method outperforms the comparison methods on both datasets.

cs.CV↗

arXiv2015

Spectroscopic Studies of an Ultraluminous Supersoft X-Ray Source in M81

Ultraluminous supersoft X-ray sources (ULSs) exhibit supersoft X-ray spectra with blackbody temperatures below 0.1 keV and bolometric luminosities above 10$^{39}$ ergs s$^{-1}$. In this Letter, we report the first optical spectroscopic observations of a ULS in M81 using the LRIS spectrograph on the Keck I telescope. The detected Balmer emission lines show a mean intrinsic velocity dispersion of 400$\pm$80 km s$^{-1}$, which is consistent with that from an accretion disk. The spectral index of the continuum on the blue side is also consistent with the multi-color disk model. The H$_α$ emission line exhibits a velocity of $\sim$180 km s$^{-1}$ relative to the local stellar environment, suggesting that this ULS may be a halo system in M81 belonging to an old population. No significant shift is found for the H$_α$ emission line between two observations separated by four nights.

astro-ph.GA↗

arXiv2015

An Updated Ultraviolet Catalog of GALEX Nearby Galaxies

The ultraviolet catalog of nearby galaxies made by \citet{Gil07} presents the integrated photometry and surface brightness profiles for 1034 nearby galaxies observed by \textit{Galaxy Evolution Explorer} (\textit{GALEX}). We provide an updated catalog of 4138 nearby galaxies based on the latest Genral Release (GR6/GR7) of \textit{GALEX}. These galaxies are selected from HyperLeda with apparent diameter larger than 1{\arcmin}. From the surface brightness profiles accurately measured with the deep NUV and FUV images, we have calculated asymptotic magnitudes, aperture (D25) magnitudes, colors, structural parameters (effective radii and concentration indices), luminosities, and effective surface brightness. Archival optical and infrared photometry from HyperLeda, 2MASS, and IRAS are also integrated into the catalog. Our parameter measurements and some analyses are consistent with those of \citet{Gil07}. The (FUV $- K$) color provides a good criterion to distinguish early and late-type galaxies, which can be improved further with the concentration indices. The IRX-$β$ relation is reformulated with our UV-selected nearby galaxies.

astro-ph.GA↗

arXiv2015

BATC 15 Band Photometry of the Open Cluster NGC 188

This paper presents CCD multicolour photometry for the old open cluster NGC 188. The observations were carried out as a part of the Beijing--Arizona--Taiwan--Connecticut Multicolour Sky Survey from 1995 February to 2008 March, using 15 intermediate-band filters covering 3000--10000 Å. By fitting the Padova theoretical isochrones to our data, the fundamental parameters of this cluster are derived: an age of $t=7.5\pm 0.5$ Gyr, a distant modulus of $(m-M)_0=11.17\pm0.08$, and a reddening of $E(B-V)=0.036\pm0.010$. The radial surface density profile of NGC 188 is obtained by star count. By fitting the King model, the structural parameters of NGC 188 are derived: a core radius of $R_{c}=3.80'$, a tidal radius of $R_{t}=44.78'$, and a concentration parameter of $C_{0}=\log(R_{t}/R_{c})=1.07$. Fitting the mass function to a power-law function $ϕ(m) \propto m^α$, the slopes of mass functions for different spatial regions are derived. We find that NGC 188 presents a slope break in the mass function. The break mass is $m_{\rm break}=0.885~M_{\odot}$. In the mass range above $m_{\rm break}$, the slope of the overall region is $α=-0.76$. The slope of the core region is $α=1.09$, and the slopes of the external regions are $α=-0.86$ and $α=-2.15$, respectively. In the mass range below $m_{\rm break}$, these slopes are $α=0.12$, $α=4.91$, $α=1.33$, and $α=-1.09$, respectively. The mass segregation in NGC 188 is reflected in the obvious variation of the slopes in different spatial regions of this cluster.

astro-ph.GA↗

arXiv2015

Combining Local Appearance and Holistic View: Dual-Source Deep Neural Networks for Human Pose Estimation

We propose a new learning-based method for estimating 2D human pose from a single image, using Dual-Source Deep Convolutional Neural Networks (DS-CNN). Recently, many methods have been developed to estimate human pose by using pose priors that are estimated from physiologically inspired graphical models or learned from a holistic perspective. In this paper, we propose to integrate both the local (body) part appearance and the holistic view of each local part for more accurate human pose estimation. Specifically, the proposed DS-CNN takes a set of image patches (category-independent object proposals for training and multi-scale sliding windows for testing) as the input and then learns the appearance of each local part by considering their holistic views in the full body. Using DS-CNN, we achieve both joint detection, which determines whether an image patch contains a body joint, and joint localization, which finds the exact location of the joint in the image patch. Finally, we develop an algorithm to combine these joint detection/localization results from all the image patches for estimating the human pose. The experimental results show the effectiveness of the proposed method by comparing to the state-of-the-art human-pose estimation methods based on pose priors that are estimated from physiologically inspired graphical models or learned from a holistic perspective.

cs.CV↗

arXiv2015

Feature Sampling Strategies for Action Recognition

Although dense local spatial-temporal features with bag-of-features representation achieve state-of-the-art performance for action recognition, the huge feature number and feature size prevent current methods from scaling up to real size problems. In this work, we investigate different types of feature sampling strategies for action recognition, namely dense sampling, uniformly random sampling and selective sampling. We propose two effective selective sampling methods using object proposal techniques. Experiments conducted on a large video dataset show that we are able to achieve better average recognition accuracy using 25% less features, through one of proposed selective sampling methods, and even remain comparable accuracy while discarding 70% features.

cs.CV↗

arXiv2014

Spectral Energy Distributions and Masses of 304 M31 Old Star Clusters

This paper presents CCD multicolor photometry for 304 old star clusters in the nearby spiral galaxy M31. Of which photometry of 55 star clusters is first obtained. The observations were carried out as a part of the Beijing--Arizona--Taiwan--Connecticut (BATC) Multicolor Sky Survey from 1995 February to 2008 March, using 15 intermediate-band filters covering 3000--10000 Å. Detailed comparisons show that our photometry is in agreement with previous measurements. Based on the ages and metallicities from Caldwell et al. and the photometric measurements here, we estimated the clusters' masses by comparing their multicolor photometry with stellar population synthesis models. The results show that the sample clusters have masses between $\sim 3\times10^4 M_\odot$ and $\sim 10^7 M_\odot$ with the peak of $\sim 4\times10^5 M_\odot$. The masses here are in good agreement with those in previous studies. Combined with the masses of young star clusters of M31 from Wang et al., we find that the peak of mass of old clusters is ten times that of young clusters.

astro-ph.GA↗

arXiv2014

Video In Sentences Out

We present a system that produces sentential descriptions of video: who did what to whom, and where and how they did it. Action class is rendered as a verb, participant objects as noun phrases, properties of those objects as adjectival modifiers in those noun phrases, spatial relations between those participants as prepositional phrases, and characteristics of the event as prepositional-phrase adjuncts and adverbial modifiers. Extracting the information needed to render these linguistic entities requires an approach to event recognition that recovers object tracks, the trackto-role assignments, and changing body posture.

cs.CV↗

arXiv2014

New 2MASS Near-infrared Photometry for Globular Clusters in M31

We present 2MASS $JHK_{\rm s}$ photometry for 913 star clusters and candidates in the field of M31, which are selected from the latest Revised Bologna Catalog of M31 globular clusters (GCs) and candidates. The photometric measurements in this paper supplement this catalog, and provide a most comprehensive and homogeneous photometric catalog for M31 GCs in the $JHK_{\rm s}$ bandpasses. In general, our photometry is consistent with previous measurements. The globular cluster luminosity function (GCLF) peaks for the confirmed GCs derived by fitting a $t_5$ distribution using maximum likelihood method are: $J_0 = 15.348_{-0.208}^{+0.206}$, $H_0 = 14.703_{-0.180}^{+0.176}$, and ${K_{\rm s}}_0 = 14.534_{-0.146}^{+0.142}$, all of which agree well with previous studies. The GCLFs are different between metal-rich (MR) and metal-poor (MP), inner and outer subpopulations, as that MP clusters are fainter than their MR counterparts, and the inner clusters are brighter than the outer ones, which confirm previous results. The NIR colors of the GC candidates are on average redder than those of the confirmed GCs, which lead to an obscure bimodal distribution of the color indices. The relation of $(V-K_{\rm s})_0$ and metallicity shows a notable departure from linearity, with a shallower slope towards the redder end. The color-magnitude diagram (CMD) and color-color diagram show that many GC candidates are located out of the evolutionary tracks, suggesting that some of them may be false M31 GC candidates. The CMD also shows that the initial mass function of M31 GCs covers a large range, and the majority of the clusters have initial masses between $10^3$ and $10^6$ $M_{\odot}$.

astro-ph.GA↗